Harmonic Measurements and Filter design
for an utility system  a case study
J. Sreedevi, K.S.Meera, B.S.Manjunath
P. Kanungao
Power Systems Division
Central Power Research Institute
Bangalore, India
[email protected]
PGCIL, NER
Shillong, India
Abstract— This paper presents the harmonic filter design study
results carried out for an electric utility for suppression of
harmonics in the system. Harmonic measurements were carried
out at 132 and 33kV substations of utility system. These
measurement results were analyzed to find the maximum voltage
and current distortions among the various measurement points.
Based on the predominant harmonics of significance harmonics
filters are designed. The filter performance results are discussed in
this paper.
Keywordscomponents: Reactive
Voltage distortion, current distortion
I.
power,
Harmonic
filters,
INTRODUCTION
Electric utilities are always concerned about having power
factor which has the advantage of reducing required equipment
ratings, line losses and voltage drops thus lessening the need
for voltage regulation equipment. This is however accompanied
by the industries increasing use of variable speed drives and
electronic equipment which can interact with power factor
correction capacitor banks to result in voltage current
amplification.
The steep development that the industries of solidstate
electronics has undergone lately resulted in the introduction of
delicate appliances which are more sensitive to the quality of
power supplied by electrical utilities in comparison with the
robust devices of the older days. At the same time these electric
appliances result in the distortion of the steady–state ac current
and voltage wave forms which results in harmonics.
The sources of harmonics are Thyristor controlled devices such
as rectifier and inverter station of HVDC schemes, Thyristor
controlled reactor, Thyristor switched capacitors and static var
compensators which introduces the harmonics into the power
circuit which results in the distorted wave forms. The other
sources of harmonics are nonlinear loads like Induction
Furnaces & Arc Furnaces, Steels mills & Rolling mills, single
Phase uncompensated railway loads and Switching equipment
& electronics etc.
Linear loads draw currents that are proportional to applied
voltages example incandescent lighting heating and motor
loads and nonlinear loads draw currents only a part of voltage
cycle and introduce harmonics. The resulting current from nonlinear loads contains 3rd, 5th, 7th Harmonics. These harmonic
currents permeate into source currents and source currents
having harmonic content impact source voltages.
Excessive harmonic voltage and/or current can cause damage to
equipment and the electrical system. IEEE Std 519199210
gives application guidelines. One of the common ways of
controlling harmonic distortion is to place a passive shunt
harmonic filter close to the harmonic producing load(s). The
harmonicproducing device can generally be viewed as a source
of harmonic current.
The objective of the harmonic filter is to shunt harmonic
current from the load into the filter, thereby reducing the
amount of harmonic current that flows into the power system.
The simplest type of shunt harmonic filter is a series
inductance, capacitance (LC) circuit which is a single tuned
filter. More complex harmonic filters may involve multiple LC
circuits.
Measurements are carried out for 9 days using Power analyzer
―PowerPro‖ instrument simultaneously at different load points
during different generating conditions. Trends in Voltage,
Current, Power and Frequency, harmonics Voltage sags and
swells can be recorded using the instrument.
The harmonic studies assume a steady state solution
corresponding to a typical operating condition. Harmonic loads
are modeled as harmonic current sources in the study. The
currents injected into the frequency dependant network are
computed for harmonic currents at each frequency and checked
for violation of limits for distortion in voltage and current. If
violation of limits is observed, the study is repeated with
appropriate tuned filters. Filters are chosen to mitigate the
harmonics, especially lower order harmonics.
II. SYSTEM CONFIGURATION
The Single line diagram of the system considered for the study
is given in Fig. 1 below. The system consists of one 132 kV
feeder from BUS1 to BUS2. There is one more feeder
connecting to BUS2, but is used only in case of emergency and
hence not considered in the study. The loads at 33 kV buses are
fed from 132 kV substation through two numbers of 10 MVA,
132/33 KV transformers. There are four feeders Feeder1 to
Feeder4 among which for Feeder 3 presently no load is
connected. Feeder1 and Feeder2 are connected to one section of
the 33 kV bus, and Feeder3 and Feeder4 are connected to the
second section of the 33 kV bus (bus coupler between section 1
and section 2 of the 33 kV bus is kept open).
The harmonic loads are represented as current sources with
sum of fundamental, 2nd harmonic, 3rd harmonic up to 16th
harmonic currents as obtained from the field harmonic currents.
The system behind the incoming feeder is represented by the
thevenin equivalent i.e as a voltage source behind short circuit
impedance. SKM Power tools – HIWAVE software is used to
carryout the filter design studies.
BUS1
1500
fashion, IEEE standard 519 1992 defines another term, the
Total Demand Distortion (TDD) as defined in
h m ax
TDD
2
Ih
h2
IL
This term is the same as the THD except that the distortion is
expressed as a percent of sum rated load current IL rather than
as a percentage of fundamental current magnitude at the
instance of measurement.
Fig 2: Current THD graph(Ia, Ib, Ic) of feeder F1
22.29 km
BUS2
TR1,132/33 kV
10 MVA
Z1=8.6%
TR2 132/33 kV
10 MVA
Z=8.6%
Feeder4
Feeder F3
BUS4
Feeder F2
Feeder F1
BUS3
630 KVA
630 KVA
CONSTANT LOADS
Fig 3: Voltage THD graph(Va, Vb, Vc) of feeder F1
(HARMONIC PRODUCING LOADS)
Measurement points
Fig 2.1: SLD of the system considered for the study
III. HARMONIC MEASUREMENTS
Harmonic measurements are recorded at 18 locations and
finally found that feeder1, feeder2 and feeder4 have harmonic
content. The parameters that are measured are RMS Voltage,
Current, frequency, Power (MW), Power factor, Apparent
Power (MVA), Reactive Power (MVAR), Total Voltage and
Current Harmonic Distortion (THD), Voltage & Current FFT,
various individual Voltage & Current harmonics, V & I Phasors
and Voltage & Current unbalance over the measurement.
Records of Maximum current THD for rated load currents and
voltage THD of Feeders F1, F2 and F4 are given in Figure 2 to
7.
THD values alone will not be sufficient to conclude the
presence of harmonics in the system. The current distortion
level can be characterized by THD values but can often be
misleading. In certain cases, there may be instances where the
THD may exhibit a high value for input current when they are
operating at a very low/light loads. This is not necessarily a
significant concern because the magnitude of harmonic current
is low, even though its relative distortion is high. To handle this
concern for characterizing harmonic current in a consistent
Fig 4: Current THD graph(Ia, Ib, Ic) of feeder F2
Fig 5: Voltage THD graph(Va, Vb, Vc) of feeder F2
Fig 6: Current THD graph(Ia, Ib, Ic) of feeder F4
substation is carried out for following cases depending on THD
levels as mentioned in previous section.
Harm
onic
order
Fig 7: Voltage THD graph(Ia, Ib, Ic) of feeder F4
This measured harmonic data is used for modeling the
harmonic loads as current sources, injecting the harmonic
currents into the network. The Maximum current THD’s and
currents RMS for F1, F2 and F4 feeders are shown in Table 1.
Table 1: Maximum THD’s of Harmonic loads
Feeder Name
F1
Current THD
Current RMS
1
2
100.00
0.375
0.0
0.0
100.00
0.307
101.6
24.7
100.000
2.229
89.7
8.9
3
1.281
42.4
0.556
30.3
1.360
60.4
4
0.140
0.0
0.381
0.0
1.307
0.0
5
3.312
158.3
0.726
176.6
1.390
75.0
6
0.202
0.0
0.174
14.4
0.824
14.4
7
0.668
0.0
0.748
91.1
1.082
129.6
8
0.213
0.0
0.216
0.0
0.192
0.0
9
0.637
0.0
0.081
0.0
0.170
0.0
10
0.149
0.0
0.074
90.9
0.513
135.5
11
2.183
0.0
0.114
53.5
2.527
86.1
Record Number
4.7
F1_2024
41.15
F1_2024
12
0.382
0.0
0.167
78.7
0.813
3.3
13
1.422
0.0
0.426
8.8
1.392
86.1
14
0.266
0.0
0.076
50.2
0.111
5.7
15
0.137
0.0
0.138
96.2
0.329
176.6
16
0.568
0.0
0.086
82.0
0.267
37.0
5.9
F4_4641
65.34
F1_1498

82.73

It is observed from the ITHD given in Table above, that current
harmonics are less on F1, F4 feeders compared to that on F2
feeder. As the harmonics are more on F2 feeder, filter design is
carried out for different cases of harmonics of F2 feeder as
given in Table 2. The detailed data extracted for F2 feeder from
the measurements is given below:
Table 2 : Highest 11th & 13th Harmonics recorded in F2 feeder
I1 (R)
F4_1498
Magnitude
Angle
IRMS
(Amp)
10.9
1
F4_4641
Magnitude Angle
Record
Number
F2
Current
F1_2024
Magnitude Angle
Max
ITHD
F4
Sl
No.
Table 3: Individual Harmonic load data for F1 and F4
Record Number
Magnitude
in kA
0.084
th
th
11
Harmonics
(Highest)
13
Harmonics
(Highest)
THD
6.06%
6.06%
10.9%
Record
Number
F2_3202
Harm
onic
order
Table 4: Individual Harmonic load data for F2
Record Number
F2_3202
Magnitude
Angle
F2_3142
Magnitude
Angle
F2_3145
Magnitude
Angle
1
100.000
28.8
100.000
82.7
100.000
85.8
2
0.448
150.9
0.26246
64.8
0.224
64.8
3
1.383
118.8
0.78266
61.7
1.037
57.7
4
0.194
17.5
0.51623
1.7
0.678
1.7
5
2.108
21.9
2.15282
35.5
1.962
29.8
6
0.099
168.5
0.12110
21.4
0.879
21.4
7
1.597
48.4
1.12586
79.3
1.001
71.2
2
I3 (B)
0.115
7.89%

10.5%
F2_3145
8
0.222
0
0.76640
0
0.787
0
3
I3 (B)
0.107

5.71%
10.3%
F2_3142
9
0.401
141.5
0.51795
0
0.476
0
10
0.245
0
0.35060
0
0.587
0
11
7.892
101.2
7.94101
105.7
7.713
105.9
12
0.871
95.2
0.73808
160.3
0.251
158.9
13
5.966
102.8
5.47117
100.3
6.022
103.1
14
0.322
67.4
0.10003
17.1
0.218
17.1
15
0.257
143.9
0.26811
36.8
0.219
36.8
0.067
3.3
0.23827
5.1
0.316
5.1
In Sl no. 1 of the above Table, maximum 11 th and 13th
harmonics are present simultaneously in recorded period. In Sl
no 2 and 3, maximum 11th harmonics and maximum 13th
harmonics are observed at different instants of time.
The Magnitude and Phase angle of individual harmonic
component currents up to 16th order harmonics of F1, F2 and
F4 loads for all the record numbers mentioned above are given
in Table 3 and 4 respectively. These individual harmonics are
considered as harmonic current sources in the simulation of
harmonic loads in the study.
III. HARMONIC FILTER DESIGN:
Harmonic Filter Design studies for arriving at the specifications
of the filter to suppress harmonics generated by Loads at utility
16
Magnitude is in % and Angle is in degrees
Two set of cases were considered for filter design studies. In
first set of cases CASE1, CASE2 and CASE3 given below,
harmonic currents at F1and F4 which resulted in highest THD
are considered. And in each case F2 harmonic load is varied as
per the highest 11th & 13th Harmonics recorded in F2 feeder as
given in Table 2.
In second set of cases CASE4, CASE5 and CASE6 given
below, harmonic currents at F1and F4 which are present during
average loading condition are considered.
Ca
se
CASE1: F1_2024, F4_4641 & F2_3202 as harmonic loads
CASE2: F1_2024, F4_4641 & F2_3142 as harmonic loads
CASE3: F1_2024, F4_4641 & F2_3145 as harmonic loads
I
II
CASE4 : F1_2024, F4_1498 & F2_3202 as harmonic loads
III
CASE5 : F1_2024, F4_1498 & F2_3142 as harmonic loads
CASE6: F1_2024, F4_1498 & F2_3145 as harmonic loads
IV
In the case of harmonic currents which resulted in highest
THD during the measurement period, highest current THD
recorded is 4.7 on FeederF1, highest current THD recorded is
5.9 and on FeederF4 and highest current THD recoded is 10.9
on Feeder F2(as given in Table 1). When the measured
individual harmonics are observed on the feeders, 5th harmonic
is dominant on F1 and 11th harmonic is dominant on F2and
F4.
Keeping in view of the dominant harmonics present, filters are
designed at the 33 kV load bus. For all cases studied, different
combinations of filters with different MVAR ratings are
considered and V_THD at all buses and I_THD of all branches
are monitored to ensure that they are within the limits as per
IEEE519. The results of the cases are tabulated giving order of
the filter and the associated MVAR rating. For each case,
I_THD on BUS1 to BUS2, V_THD at BUS1, BUS2, BUS3 &
BUS4 are tabulated in Table 5. The 33 kV bus voltages at
substation are also monitored to make sure that there are no
overvoltages with connection of filters. To have an idea of the
possible rise in voltage at 33 kV bus of substation with
connection of filters, 33 kV bus voltages is also tabulated for
different cases studied.
V
VI
Harmonic
filters
Table 5: Results for different cases
I_TH
V_THD
D
33 kV
voltage
Orde
r
MVAR
BUS1
to
BUS2
BUS1
BUS2
BUS3
BUS4
BUS3
11
11
11
11
11
11
11, 5
11, 5
11, 5
11, 5
11, 5
11, 5
2
1.5
2
2
2
5
2, 1
2, 1.5
2, 1.5
2,4
2,4
2,5
5.22
2.39
2.51
5.47
2.30
5.30
2.37
5.5
2.97
3.37
2.52
2.36
6.45
2.62
2.48
6.39
2.61
2.50
0.56
0.20
0.23
0.61
0.20
0.57
0.20
0.57
0.21
0.2
0.19
0.19
0.70
0.21
0.2
0.67
0.2
0.2
0.92
0.32
0.36
1.00
0.32
0.93
0.31
0.93
0.33
0.31
0.30
0.29
1.16
0.34
0.32
1.10
0.32
0.31
0.97
0.34
0.38
1.05
0.33
0.98
0.33
0.97
0.35
0.32
0.32
0.31
1.21
0.35
0.32
1.15
0.32
0.31
0.92
0.32
0.36
1.00
0.32
0.93
0.32
0.93
0.33
0.31
0.30
0.29
1.16
0.34
0.32
1.11
0.32
0.31
32.06
32.14
32.12
32.00
32.08
32.04
32.12
32.05
32.14
32.27
32.18
32.20
31.99
32.14
32.26
32.04
32.30
32.39
Fig 8: Distortion Voltage Spectrum at
BUS3 and BUS1 without filter
Fig 9: Distortion Voltage Spectrum at
BUS3 and BUS1 with filter
Fig 10: Impedance Magnitude Scan at
BUS3 and BUS1 without filter
Fig 11: Impedance Magnitude Scan
at BUS3 and BUS1 with filter
Fig 12: Impedance Angle Scan at
BUS3 and BUS1 without filter
Fig 13: Impedance Angle Scan at
BUS3 and BUS1 with filter
From the simulations carried out it is observed that 11 th
harmonic filter of 2 MVAR is sufficient to keep the harmonics
within limits specified in IEEE519 for CASE1, CASE2 and
CASE3. Whereas for CASE4, CASE5 and CASE6, it is seen
that in addition to 11th harmonic filter, a 5th harmonic filter of 5
MVAR is required.
The Distortion Spectrum of harmonics and Impedance
Magnitude Scan and Angle for CASE 1 are provided in Fig 8 to
Fig 13. Distortion Spectrum of harmonics at 132kV bus and
33kbus without filters is given in Fig 8 and with filters is given
in Fig 9. The presence of 11th and 13th harmonic magnitudes is
seen in distortion spectrum without filter where as in distortion
spectrum with filters they are absent. Impedance Magnitude
Scan and Angle at 132kV bus and 33kbus without filters is
given in Fig 10 and Fig 11 respectively and with filters is given
in Fig 12 and Fig 13. BUS0006 and BUS0001 in the figures
refer to BUS3 and BUS1 of single line diagram.
However, since the filter design study considers the presence of
maximum harmonics simultaneously in the system, which may
not occur in practice the 11th harmonic filter of 2 MVAR be
implemented in the first stage. Later, in the next stage the 5 th
harmonic filter may be considered for implementation after
confirming the modified THD levels (after implementation of
11th harmonic filter) in the system.
In Table 7, Filter spectrum Report is provided which gives the
fundamental and harmonic currents through the filter
components.
Table 7: Filter spectrum Report
Additionally, as the source of the 11th and 13th harmonics is on
feederF2 the suggested 11th harmonic filter can be installed at
the 33 kV load bus itself so that the filter can be switched in/out
along with the load.
V. RESULTS & DISCUSSION:
5.1 Specifications of 11th harmonic filter:
From the above studies specifications of 11th harmonic filter are
given in Table 6 below:
Table: 6: Specifications of filter
Sl.No.
Parameter
1
2
3
4
5
6
7
8
9
Filter Type
3 Phase MVAR rating
Rated Voltage
Rated frequency
Quality factor
Connection
Resistor(R)
Inductor(L)
Capacitor (C)
Single tuned filter
2.0 MVAR
32.417 kV rms
50 Hz
20
Y
2.5781
0.0143 H
5.8459 F
The 2MVAR, 11th harmonic single tuned filter designed will
act as a 2MVAR capacitor at fundamental frequency of 50Hz.
The inductance and the quality factor, Q (wL/R at tuning
frequency) for the harmonic filter reactor is specified at the
resonant (tuned) frequency of the harmonic filter network. The
tolerance on the inductance should be selected to ensure the
proper performance of the harmonic filter network across the
range of the tolerance. Inductance tolerance is usually available
as ± 3% or +0% to 6%. The tolerance on the quality factor, Q,
is typically ± 20%
IEEE Std 182002 states that, capacitors are intended to be
operated at or below their rated voltage. Capacitors shall be
capable of continuous operation under contingency system and
bank conditions provided that none of the following limitations
are exceeded:
a) 110% of rated rms [rootmeansquare] voltage
b) 120% of rated peak voltage including harmonics, but
excluding transients
c) 135% of nominal rms current based on rated KVAR and
rated voltage
d) 135% of rated KVAR.
Harmonic filters are generally not tuned to exact harmonic
frequencies, as tuning directly to the harmonic frequency may
have the following undesirable consequences:
a)The low impedance at resonance can result in nearly all
harmonic current at that frequency being absorbed by the
harmonic filter. The harmonic filter is required to be larger and
more expensive than is needed to achieve the required
harmonic performance
b) The harmonic filter interaction with the system impedances
results in a parallel resonance at a frequency just lower than the
tuned frequency. If a harmonic filter is designed exactly at the
harmonic frequency, a variation in the impedance values of the
actual equipment from the design values could retune the
harmonic filter and place the parallel resonant frequency very
close to the harmonic. Instead of low impedance, the combined
harmonic filtersystem impedance becomes resonant at the
harmonic frequency, distortion levels become unacceptable,
and damaging voltage amplification may result in severe cases.
Changes in the system supply can shift the parallel resonance
also.
In this regard, literature survey indicated either 0.96 or 0.98
detuning factor being adopted. In this study detuning factors of
0.99, 0.98 and 0.97 are considered and CASE1 presented in
Table 5 is repeated with filter harmonic orders of 10.89, 10.78
and 10.76. The results of this study, considering detuning
factors is presented in Table 8.
From these results it is seen that detuning factor 0.98 results in
I_THD of less than 2.5, and detuning factor of 0.97 results in
I_THD of 2.55 which is violating the limits.
Table 8: Results for detuning frequencies
Harmonic
I_TH
V_THD
filters
D
Order
M
BUS1
BUS
BUS
BUS
BUS
VA
to
1
2
3
4
R
BUS2
10.89
2
2.43
0.21
0.33
0.35
0.33
10.78
2
2.48
0.22
0.34
0.36
0.34
10.67
2
2.55
0.23
0.36
0.37
0.35
Cas
e
I
33 kV
voltage
BUS3
32.143
32.143
32.143
Table: 9: Specifications for detuned filter
Sl.No.
Parameter
1
2
3
4
5
6
7
8
9
Filter Type
3 Phase MVAR rating
Rated Voltage
Rated frequency
Quality factor
Connection
Resistor(R)
Inductor(L)
Capacitor (C)
Single tuned filter
2.0 MVAR
32.417 kV rms
50 Hz
20
Y
2.5255
0.0149 H
5.8459 F
From the studies carried out 11th Harmonic filter is
recommended to mitigate the harmonics as per IEEE 519 guide
lines. A detuning factor of 0.98 is to be adopted for the 11th
harmonic filter. The Filter may comprise of two similar star
connected capacitor banks in parallel with the a CT on the tie
between the two star points.
ACKNOWLEDGEMENTS
Authors wish to thank, CPRI for permitting to publish this
paper. We also wish to thank Dr. R.S. Shivakumara Aradhya
and Sri.R. Deshpande for their support.
VI. REFERENCES
1.
2.
3.
―Report on Harmonic measurement in Powergrid, NE
region, April – 2010‖ CPRI report: DCCD11341
dated 2241010.
IEEE5191992: Recommended Practices and
Requirements for Harmonic Control in Electrical
Power Systems.
IEEE Std 18.2002, IEEE Standard for Shunt Power
Capacitors.